Harmful generation in LLMs relies on a compact, unified set of weights that alignment compresses and that are distinct from benign capabilities, explaining emergent misalignment.
If no configuration meets this threshold, we select the one with the lowest StrongRE- JECT score among configurations whose utility remains within 95% of the unpruned model
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Large Language Models Generate Harmful Content Using a Distinct, Unified Mechanism
Harmful generation in LLMs relies on a compact, unified set of weights that alignment compresses and that are distinct from benign capabilities, explaining emergent misalignment.